Usage

Clone this repository and download the required anaconda packages Start in the src directory

To run a file, use the command python [file] [options]

Where file is the name of the file to run the analysis with, such as occupancy_clustering.py, and options are the command line args.

Runnable Files

landsat_classification.py - Logistic regression model landsat_deep_learning.py - Deep learning model occupancy_clustering.py - K-means and GMM traffic_bayesian_regression.py - Bayesian Ridge Regression

Args

--test TEST_FILE path to additional features to test against. Path must be relative to the current directory. If supplied, results of predictions against this test data will be the last thing printed by this script (optional) --re-train will re-train the model and document cross validation process --analysis will run the model against test data and report performance --test-dev Will do a train-test and print the results to the command line using the dataset from Canvas. For development purpose only

Train

--re-train Overrides the existing model with a new one Inputs: none Outputs

• Model file saved in /model dir
• CV results saved in /cv_results dir

Test

--test TEST_FILE Makes predictions on new unseen data Inputs: file with new data (may or may not have target column?) Outputs: file with predictions in /results dir

Performance

--analysis Conducts performance analysis on the trained model Inputs: none Outputs - Print accuracy/error/variance to console - File with untuned and tuned performance scores - Performance results in /performance dir